Thanks for your response, Debbie. I've found several Mary's who are also called Polly in those days, so you're likely right. Must've been a cultural thing of the times.
Yeah, George is quite a question mark for me right now. I recently had him married to Mary Denslow in Jennings County based on a marriage record that looked soooooo much like "...low" in both names as matches. But I couldn't justify why he would go to Jennings county to marry her; and another researcher pointed out that there was a George Harlan with a Mary living in Jennings County the next census (age of child in line with a new marriage), citing that same marriage record as documentation. That made sense. So I deleted my data. When I did, some DNA matches disappeared. Hmmmm.... But I also deleted a line I'd added based on a similar situation at about that same time, so I'm not certain which line had the connection.
So, did you DNA test and, if so, do you have someone named Waugh in your undocumented DNA matches? If so, I may need to rename Mary Denslow as George's wife and my ancestor.
I haven't found anything of Mary "Polly" Kennedy except a Kennedy living nearby on one census. And here's the clincher. My mother said that my grandparent was 1/4 or 1/8 Indian, which would fall in George's line. A full-blood Cherokee friend told me once that the best way to know you have Indian blood is when you can't find records on your ancestor. So, have we found the Native American blood my Mom always swore by in George? Or is he just missing at the moment?
Let's keep in touch, shall we?